JunoCam "Marble Movie", July 10 through October 14 Options

[tedstryk]

Selected imgages showing Jupiter's north polar region:
[attachment=40038:JNCE_201...0_sqr_96.jpg]
This is somewhat quick and dirty, but I first wanted to submit (the png version of) this overview to the missionjuno site, before spending too much time with remaining issues.
Btw.: Three of the rgb images of part 7 are dedicated to (a?) moon(s).

Great work. Where did you get the data for these?

[Gerald]

I've completed the upload of part 7 of Marble Movie level 1 stills (about 250 MB zip file).
Three files are not included: image #6080 appears to be corrupted, images #6123 and #6151 confused my Jupiter detection algorithm for some as-of-yet unknown reason, and selected a moon instead as object of interest.
If I find the reason for the confusion, I'll add these two images later. The last image of part 7 is cropped in the processed version; I may provide a completed version later.
Note, that the zip contains two series of images of different naming "003" and "005" infix with different size and supersampling.

Raws are from the missionjuno site. Anything else I've derived from these raws.

[JRehling]

Here's a little elaboration on the largest image in Gerald's montage: I divided the Jupiter image from a Lambertian-shaded sphere at approximately the same phase as Jupiter. This allows for more detail to be seen in a band several degrees wide near the limb. This information is essentially redundant at lower latitudes (we could wait and see what's in the dark to rotate into light), but it gives us a uniquely clear vision of the immediate vicinity of the pole. (I've been applying this transformation helpfully to my own images of Mars and Mercury in order to map them.)

The swirling storms in the polar region seem almost randomly distributed, the longitudinal patterns that dominate elsewhere almost absent here. Potentially a few hours' images could give us a clear vision of the entire polar region, but for now, this is just the one snapshot.

Attached thumbnail(s)

 

[ZLD]

Hadn't realized how the data was be distributed until today. Had a go at one of the images. Workflow seems quite long on assembling these. Have I missed a tool that makes this easier?

[Ant103]

I spend a good part of the night, and the middle of the day trying to figure out how getting a good result with the latest JunoCam images from Marble Movie 7, and didn't get a good result. Because Juno is moving along her orbit toward Jupiter, so we have a consequent shifts between two sets of RGB scan. Maybe with complex transformation, reprojection, but for now, I will not work with RGB layers, only the red ones.

Here's an enhanced version of one red layer, at least

Attached thumbnail(s)

 

[mcaplinger]

I spend a good part of the night, and the middle of the day trying to figure out how getting a good result with the latest JunoCam images...

Somehow that sounds a lot like my recent experience.

BTW, if people were unaware, Cassini took excellent movies of the polar regions during its Jupiter flyby back in 2000. Well worth reviewing as you're waiting for Juno data. http://www.ciclops.org/view/81/Jupiter-Polar-Winds?js=1

[elakdawalla]

Gerald's got quite a head start on the rest of us

I've now updated my Marble Movie data page with Gerald's latest thumbnails, and added links to his PNGs. The raws are beyond my capability to process so I look forward to the efforts of others here!

[mcaplinger]

Hadn't realized how the data was be distributed until today... Workflow seems quite long on assembling these.

And we transition from complaints about not having the data to complaints that the data are hard to work with.

The marble movie is a bit of a special case. Because of the large number of typically tiny images, it was decided to not use the normal missionjuno "gallery" mechanism but to just release them as a large unprocessed blob. Unfortunately the last batch of images of the marble movie was close enough to the planet to be interesting but didn't make the cut to be included in the gallery. The gallery images will be available in both raw and a couple of different processed forms and in general be easier to work with.

I don't know of any public tools to work with the raw data. These could be as simple as a few lines of some scripting language to rearrange the filter strips to a version of something like the ISIS support for MARCI or LROC WAC suitably modified for Junocam.

[ZLD]

Emily, I notice in the final image on your page, your thumbnail label says "no blue" but the registered label for the raw says it is an RGB capture - unless I'm reading the label wrong. Good resource as always, nonetheless.

mcaplinger: Not complaining that the data is challenging, just time consuming. Also, I already hammered out several scripts to cut the time down significantly. Most certainly a necessity if you plan to do more than a few of these.

[Ant103]

This is the best I could do with the raw image I was working on since yesternight. I chosen to use the greyscale red layer as luminance channel on the RGB mosaic, blurred a little to hide the artifacts.



Enchanced and color balanced version :



http://www.db-prods.net/blog/2016/09/01/qu...-junocam-image/

I'm very curious about the imagery taken much more closer to Jupiter, and pretty afraid by the amount of work waiting for us…

[elakdawalla]

Sorry, that text was confusing. It referred to the fact that they didn't take a blue-only frame at that time. I've changed the text to be clearer.

[Gerald]

This 1-second fragment of part 7 of Marble Movie level 2, is sufficiently small, that I can post it here directly:
 juno_marble_jupiter_7b.avi ( 350.56K ) Number of downloads: 305

The sequence is immediately before the larger Jupiter images begin.
I've supersampled the latter with 60°x60° and 120 pixels per degree in order to stay continuous with the small marble images. This drove my computer to the limits, and required about 8 minutes per frame.

Besides a few corrupted frames, level-2 Marble Movie parts 1 to 7 is completed. I'll upload this version in a few minutes as an AVI; it will show up on this site in half an hour or so. EDIT: Upload completed, you may need to refresh your browser window.
I'll try to fix the remaining issues tomorrow.

Re tools: Learning means re-inventing. So I've written the whole JunoCam processing from scratch in C++, including developing the required math as far as it's not standard graduate, but except graphics file format conversion. Similar software may or may not exist elsewhere.
In the meanwhile, it's about 50 executables, and several MB of source code. This includes analysis and calibration software.
I'm expecting another two dozens or so tools still to be written to cover all calibration and processing I'm intending to perform with JunoCam images.
The software will likely not be released, in order to avoid multiple issues, like possibly conflicting professional obligations, legal and security issues, or documentation and support overhead.

However, I'll release processed JunoCam image products public domain as far as possible.

[xflare]

I've now updated my Marble Movie data page with Gerald's latest thumbnails, and added links to his PNGs. The raws are beyond my capability to process so I look forward to the efforts of others here!

If you let all the images on that page load and then scroll up and down it superfast - it turns into a kind of zoetrope movie !

[Roman Tkachenko]

An attempt to create a polar view.

[Gerald]

A successful attempt!

---

Repaired (filled missing part with black to get full framelet triples), and level-1-processed image #6080:


I also found and fixed the software bug leading to occasional wrong planet identification and crop of the last part 7 Marble Movie image.
RGB Images #6068 to #6159 are undergoing reprocessing for the AVI, with the according CPU marathon over several hours.